How to Make Sure Brands Can Activate Their Identity Data
A lot of brands have reached a turning point with their first party data. Many have already done the heavy lifting to aggregate internal data, and some have even created their own internal ID to act as a single source of the truth for each customer or prospect.
Now many of these brands are hoping to connect their data to the larger data ecosystem to activate audience-targeting and measurement. The two most important things to consider are that data is kept secure through the process, and that the outcome is as effective as possible.
There are a lot of different ways to get this done, with partners like AWS and Snowflake, brands can essentially park their data and match it to anyone else working with these services. Or brands can go direct to publishers or other data owners and match their data using specific data clean rooms. In each scenario, brands need to consider if and how their data might be exposed, and how to maximize their ability to reach their audience.
Ensuring Data Security
When an ID is ready to be used externally, the first order of business is to ensure the data is shared in a secure environment, in a place like AWS, Snowflake or in a data clean room. We are reaching a point where sharing data in these clean environments is becoming much more common. Places like Snowflake and AWS allow data to be joined in such a way that data doesn't need to be moved (and therefore potentially exposed).
However, it's highly likely that every brand will end up having to work with partners who have their data somewhere different. As soon as a brand needs to move data, there is a security risk. Data could be exposed broadly and stolen by a hacker, or it could be exposed to a partner, allowing the partner to copy information and use it for their own benefit.
To ensure that data is not exposed broadly, brands need to invest in creating a secure workflow for sharing encrypted data. This could mean working with randomized keys or investing in multi-party computation.
Another major concern is that brands don't want to expose their data so that their partner can use it. If a large advertiser has a valuable audience that they want to match to a publisher, that advertiser doesn't want that publisher to access their data or gain key insights from that data matching.
This is where a data spine or ID graph comes in. Brands can add in a third party that has one to create noise to eliminate the risk of exposure. Clean rooms allow two companies to connect internal data, whereas an ID graph can create enough noise so that neither party could determine exactly how everything is connected.
Activating Data for Advertising and Measurement
There are also several different ways brands can get their ID into the ecosystem in order to activate targeted ad campaigns. There are some brands that will map to an ID like The Trade Desk's UID 2.0, and others that create a private marketplace to share the data.
The benefit of a private marketplace is that it is most secure. A brand can control who they are sharing the data with, how they share data and how rare using the data. This creates a walled garden for marketers to and publishers collaborate, build audiences, and activate.
Alternately, appending a universal ID, there may put data at a higher risk. That UID could be matched back to data like email. To combat the risk, The Trade Desk is working to refresh IDs on a high cadence. This solution creates operational challenges because data must be refreshed in the platforms being used, which can become an operational burden as brands try to combine insights over time or across partners.
Maximizing Value
With such high-fidelity data, consented first party data, the scale will be lower but should yield better results. With that said, match rates are a major consideration for brands as they look for partners that can help them activate as much of their data as possible. This is where large partners with robust identity graphs can be valuable.
If a brand tries to connect directly, match rates will be low. A third party with a fully built out identity graph will increase match rates and scalability. However, not all identity graphs are created equal. Brands should test a variety of identity graph partners to see who provides the right results, as everyone uses different foundational data and a proprietary methodology.
The only way to know is to test. The next year will be one of testing and learning as brands determine who their best partners are, and companies build out pipes to ensure secure and effective data matching. While all this testing is good, we all need to remember what we're trading in. First party data is a high stakes game, and our industry could be penalized if we don't adhere to best practices.
The views and opinions expressed are solely those of the contributor and do not necessarily reflect the official position of the ANA or imply endorsement from the ANA.
Nancy Marzouk is the CEO and founder of MediaWallah.